Agricultural development is facing the dual challenges of intensifying natural resource constraints and global warming, making the development of low-carbon agriculture imperative. We adopted a meta-analytic framework integrating multilevel meta-analysis, mixed-effects meta-regression, and random forest analysis to investigate the overall effect sizes of agronomic measures and environmental factors on the carbon footprint intensity (CFI) within the narrow-sense crop production boundary. Using 181 independent effect sizes, the overall evidence indicated an average CFI reduction of 14.11% across the examined factors. In the multi-moderator model, organic material return (OMR) was identified as a key moderator. In particular, returning plant-based organic inputs and applying biochar were associated with marked decreases in CFI, by about 22.60% and 49.50%, respectively. Conservation tillage (CT) showed a decreasing trend (23.40%), but this was not statistically significant overall. The synergistic combination of OMR and CT, two advantageous agronomic measures, significantly decreased CFI by 32.66%. Random forest analysis provided a clue interpretation regarding the regulation of the environment (R² = 0.0811). Temperature and precipitation exhibited regulatory potential, and soil pH also showed notable importance. Three-way interactions among soil pH/temperature, OMR, and CT were not significant, whereas precipitation showed a marginally significant three-way interaction with OMR and CT, suggesting that the effect of OMR on CFI depended on both precipitation amount and CT regime. This study not only assessed factors contributing to the carbon footprint of crop production but also identified interactive effects between advantageous agronomic measures and environmental factors, providing theoretical support for the precise management of green, low-carbon agriculture.
Tao et al. (Fri,) studied this question.